Gold: What We’re Up To…

Development of Geegeez Gold into a form provision for all punters, regardless of time, experience or desired level of engagement, is ongoing. From the outset, we've sought to differentiate from other form books by using more visual indicators, and aggregating data for expediency purposes. Nowhere is this more obvious than in the Instant Expert view.

 

More configurability is now yours with Instant Expert v2.0

More configurability is now yours with Instant Expert v2.0

 

We recently gave Instant Expert its first nip and tuck - more collagen injection than full cosmetic surgery - by introducing additional filters for time period, race code, and handicaps/all races. And we continue to make progress with the Query Tool, though that has been a little slower than ideal.

 

Basic charting functionality in geegeez.co.uk Query Tool

Basic charting functionality in geegeez.co.uk Query Tool

 

The pipeline for Gold development has a long route to travel, flowing as it does from the contents of the darker crevasses of my creaking cranium, via a couple of wildly talented but somewhat maverick developers, into the estuary of our test site, and finally onto the live website here.

We use a few collaboration tools to manage that flow, and in the remainder of this post I will share some of what currently resides there.

But first a warning...

Before I share what next, I need to warn you of the very next thing we're undertaking.

As Geegeez Gold has grown, both in terms of functionality and user numbers, so the requirement for computing power has grown too. Also, time has passed since we kicked this off in 2013, and the server we were using then is no longer up to the job.

We need http2, and fast-cgi, and SSD, and TLS 2.1 and other acronyms that I don't fully understand but have on good authority will enhance both the stability of the site - which, in truth, has been disappointing in the past month or so - and the speed of access for users. Oh, and of course, the new box will further improve security.

It's a hideous job, fraught with peril and, to be blunt, I've been sh!tting myself about doing it for more than two years now. But it cannot be ignored any longer. We are currently in the process of building the new server - a process that has complexity relating to the various inputs we have, as well as code and data challenges, etc etc blah blah.

From your perspective, you just need it to work and, in an ideal world, to work a little faster. Our job is to make that happen. We're on it, and we'll likely be switching in the next fortnight or so. *dons tin hat, and assumes the brace position*

 

Then what?

Once we've migrated to the new whizzbang (more whizz less bang, I hope) server, we can get back to the fun stuff. Here's what's in store, though keep in mind that not all of it will necessarily see the light of day: the features we're considering range from 'must have' to Coleridgesque opiate-addled fantasy!

Here we go then...

- Calculate and publish overrounds for races (general excitement level: yawn)

- QT: make 'group by' links clickable (general excitement level: useful)

- QT: add 'tool tips' to explain stuff (general excitement level: yawn)

- QT: add Speed Rating Rank to allow analysis of the effectiveness of the figures (general excitement level: cool)

- QT: add damsire (general excitement level: mildly interesting)

- QT: allow users to save queries within QT (general excitement level: awesome!)

- QT: allow users to view saved query qualifiers on the racecard (general excitement level: totes amazeballs!)

- QT: add Equipment Count (e.g. blinkers first time) (general excitement level: interesting)

- QT: add Wind Count (e.g. first run after wind op) (general excitement level: mucho interesting)

- QT: add last run info (days since, class, course, distance, etc) (general excitement level: very useful)

- Then What?: add class distinction to 'then what?' form follow in racecard (general excitement level: useful)

- Top SR Differential report: highlighting the horses furthest clear of their fields each day (general excitement level: interesting)

- Export report output to Excel (general excitement level: nice touch)

- TJ Combo report: add one year trainer/jockey sub-report (general excitement level: very useful)

- Class Move report: new report, with an option to display class move indicators on racecard (general excitement level: useful)

- Racecard menu filters: only interested in sprint handicaps? See only the races you're fussed about (general excitement level: niche)

- Instant Expert: add draw position (general excitement level: very useful for flat races)

- Instant Expert: create trainer/jockey/sire options, if possible (general excitement level: awesome)

- Ratings pars: calculate and publish ratings pars for course/distance/class combinations (general excitement level: interesting)

 

And now for something completely different...

Lots of interesting bits there, most of them of interest to Peter but not Paula, or Paula but not Peter, if you see what I mean. Alongside all of that, I am working on a couple of more mechanical features, which are, to be frank, no better than even money to see publication. For interest, then, these are they...

  1. Sprint Handicap Form Cheat Sheet

I recently read an old book called Betting on Flat Handicaps, by Jon Gibby. It was published fifteen years ago and, since then, much of the draw and pace data in the book has become outdated. Luckily for us, Geegeez Gold has permanently current data for both of those elements which, applied to the excellent regimen espoused in Gibby's book, provides a very solid framework for form study in such races.

Nothing about his approach is earth-shattering or rocket science, but all of it represents common sense fundamental principles. Gibby, who focuses primarily in sprint handicaps, goes about eliminating a section of the field based on where they are drawn. For the remaining runners he digs into the form book. See, I told you it wasn't rocket science. That it is effective should not be surprising.

I've been looking at a race or two a day, when I have time, and my results in this 'testing phase' are below.

Very (very!) early days, but a good start

Very (very!) early days, but a good start

 

Now, straight off the bat it needs saying that a 40% strike rate is unsustainable, still less when two of those winners were at double digit odds. So let's be realistic here: there's a solid chance of a protracted losing sequence coming next for the test. What I'm looking for in this incubator phase is some sort of affirmation for the process. I want the picks to largely run well, or with legitimate excuses; and I want unbacked winners to generally have been on my shortlist.

If after 50 bets - still a small sample but data-driven betting on racing is usually based on small samples - that's the general feel I'm getting, it'll be time to engage a little more committedly...

Your first 30 days for just £1

So, if the above is the answer - or are the answers - so far, where are the workings out?

My process is as follows:

  1. Draw
  2. Trainer form
  3. Pace profile
  4. Horse form

Here's how I go about it:

Draw

For the draw, I go to the draw tab in the race card on Geegeez Gold, and select the going above and below today's official going (e.g. if today's going is good to soft, I'll choose the range good through to soft). I then amend the runner range to be +/- 2 (e.g. if ten runners, I'll select 8 to 12). These ranges are used to get better sample sizes from which to work. Clearly, by being less specific on going and field size I lose a little in terms of direct relevance; but my contention is that this is more than mitigated by the slightly higher confidence levels commensurate with a bigger dataset. I also use 'actual draw', i.e. number of stalls from box 1 after non-runners are considered. Anyhoo...

I then have some data in the IV column on the draw tab. IV, or impact value, is a measure of how much more or less likely something is to happen. In this case, it's a measure of how much more or less likely a horse is to win when drawn in stall x based on the total population of runners in the sample.

I add that IV number to my spreadie and then, in an attempt to even out the individual starting stall IV's, I take an average of the stall and its immediate neighbours. That is, for stall 1, an average of stall 1 and 2; for stall 2, an average of stalls 1 to 3; for stall 6, an average of stalls 5 to 7; and so on. To that, I add a little crass colour coding (what can I say, I find colour a very powerful visual aid...!)

It looks a bit like this.

Colour-coded average draw IV's

Colour-coded average draw IV's

 

The colour-coding is a bit rough and ready, and I'm obviously trying not to cherry-pick. That is, it makes no sense, for instance, to infer a positive impact from horses in stalls four and six, but negative impact for the horse in the stall between them. We need to be a little 'real world' here and look for general patterns. The averaging thing helps to some degree, but there remains inference in the process.

So, in the example above, I have a primary draw focus on those in stalls one to five. Note how the middle draws have fared less well historically in this example. The winner here was drawn five and, if I'd looked solely at individual draw, stall five's IV of just 0.42 would have put me off. Even in this case, it remains daft to say that five is green/good while six is red/bad. With the exception of some tracks where stall 1 is a negative, it is generally the case that there is a steady diffusion of goodness/badness rather than absolutely/arbitrary cutoffs. But this is punting, and we need to take a view!

 

Trainer Form

I'm not certain I've got the best approach to evaluating draw, and that's a comment which applies even more to my current method for trainer form. Here, I'm taking an average of the IV's for each trainer's 14 day, 30 day, course 1 year and course 5 year form. The problem here is that there is a very obvious 'related contingency': in plain English, the trainer's 30 day form includes his 14 day form, and the trainer's five year course form includes her one year course form. So that's probably wrong. It might be better to add the average of the 14/30 to the average of the C1/C5, or to take an average of those two averages.

Moreover, I'm not currently factoring in the contextual form elements, such as trainer's form with e.g. sprinters, or first run off a layoff or handicap first time, etc. That, too, is probably wonky.

Nevertheless, I have some data, which gives me a view of trainer form, and looks like this:

An approximation of trainer form is useful ballast

An approximation of trainer form is useful ballast

 

Pace Profile

Step 3 is to profile the pace in the race. How much is there overall, how is it spread across the field, and what are the individual horse's pace profiles?

All of this can be gleaned very quickly by sorting the runners on the pace tab by draw. Here is how our example race looked on the pace tab:

Pace tab shows overall pace profile, historical performance, and individual pace preferences

Pace tab shows overall pace profile, historical performance, and individual pace preferences

 

There is a lot of information in this view.

First, in the green box, I've highlighted the historical pace performance of the different early run styles - with the going and runner ranges extended as discussed previously.

Second, in the blue box to the right, I've highlighted the runners' average pace score for their last four UK/Irish runs (4 - led, through to 1 - held up).

And third, the view is sorted by draw (see left hand orange box) to give a visual perspective on how the pace might play out, based on the recent history of the runners in the field. In this case, there looked to be a bit of pace competition both high and low, where our draw research ssuggested those drawn on the flanks had the best record. Moreover, those who raced prominently or mid-division had fared marginally best in this sample (see the coloured blobs in the highlighted green box), though there seemed little in it.

That draw/pace profile hinted at Muscika, Black Isle Boy and My Name Is Rio.

 

Horse Form

The final piece in the puzzle is horse form. Clearly, this is a significant piece and, once the field has been whittled using draw and/or pace, the focus is keenly on what the animals have done. Trainer form is probably the least considered element at this stage in my testing, unless a handler is obviously bang out of, or in, form. [Trainer form is also the element which needs most further scrutiny just now, in terms of how I measure it].

I use Instant Expert, Full Form and the Card tab for the horse form part, and it feels like the one component which will be quite difficult to automate.

Here's how the key components of my 'race card' looked in the spreadsheet [click the image to expand it].

Pulling all the elements together offers a pretty solid understanding of which horses have value chances

Pulling all the elements together offers a pretty solid understanding of which horses have value chances

 

Looking back on this I probably got the pre-runner pace profile colouring wrong. It is generally my preference to favour early pace, as I  have done here; but the historical profile suggested a slightly more restrained ride was often the way to go, and so it proved. Fortunately, there was enough in Black Isle Boy's favour - especially at the price - to have a small interest anyway, though I feel I was somewhat lucky rather than good in this instance.

 

Why are you telling me this?

I mention all of the above for two reasons. One, a bit like the 'secret' to weight loss being diet and exercise, there is no secret to form study. It involves pulling all, or as many as possible, of the salient factors into a melting pot of deliberation. It takes time and effort, some of which can be automated.

Which segues nicely onto reason two: all of this content exists within Geegeez Gold, which means the process of automation is within our/my control. It's on the wish list!

 

2. A R-r-r-r-r-ating?!

Ratings are tricky. Super tricky. There are times and places when they're of huge utility, and there are other times when, in my opinion, they offer  little to no value. For example, what use is a speed rating if the race looks likely to be run at a crawl? And what use is an ability rating if it fails to account for the specific race conditions on the day?

The best rating would accommodate such considerations and more besides; and it would learn to refine its number set as more evidence is presented. In point of fact, that's largely how Peter May's ratings, which we proudly carry here on Geegeez Gold, are derived.

And, in a weak moment, I got it into my noggin that I might create my own set of figures. Actually, it's a recurring thought. But I know that this would become somewhat sisyphean (I'd love to name a horse, Somewhat Sysiphean!) - it would be a life's work, and one almost certainly doomed to ignominious failure.

In spite of holding that contention in my saner moments, I did draw up the first (extremely) rough draft of 'ground zero' for a rating set. Here is it. Click for a full size view, and feel free to make a comment below. But please don't ask questions!!

Some 'fag packet' thoughts on how a rating set might be constructed. Not. Straightforward.

Some 'fag packet' thoughts on how a rating set might be constructed. Not. Straightforward.

 

Summary

So yes, oodles in the pipeline, some of it nearer / more realistic than others; all of it obliged to follow on from the major infrastructure work we're currently undertaking. It's going to be another exciting year ahead!

Matt

Instant Expert v2.0 is LIVE

It's live, the new Instant Expert v2.0. Or maybe we'll just continue to call it Instant Expert, eh?

Most importantly, if you're in the Remain camp, do nothing and Instant Expert will continue to display the data as ever it did. However, if you're an Instantexpiteer (see what I did there? Not great, granted) then you'll want to have a watch and a listen to the below videotape, which explains all...

There is also an updated User Guide that outlines the changes. You can get that from the link on the My Geegeez page.

Geegeez Cards: A few small tweaks

Nothing exciting in the latest tweaks, which are little more than minor bug fixes. Specifically, they're as follows:

- Disabled switching between race card tabs when using the arrow keys for up/down scrolling

- Added 'abandonment' notifications for races/meetings

- Fixed a bug with scrolling between Full Form runners from the drop down, whereby performance slowed the more races that were viewed

- Added more race details (distance, runners) to compact card menu 'hover over' data

- Fixed issue with incorrect 'runs since wind surgery' counts

More exciting things coming soon...

 

Part 2: The Importance of Pace in 5f Handicaps

In my first article I looked at pace in 5-furlong handicaps focusing on the running style bias angle. The figures clearly showed a huge difference between the front running chances of horses depending on which 5f course he/she was running. In this second part, we will revisit the course angle and aim to offer a more complete picture.

To recap from the first article, when I talk about pace my main focus is the early pace in a race and the position horses take up early on. The Geegeez website splits pace data into four groups - Led, Prominent, Mid Division and Held Up. These groups are assigned numerical values – led gets 4 points, prominent 3, mid division 2 and held up 1. On each Geegeez racecard these figures are assigned to every horse in the race going back four UK or Irish runs.

We can use these numerical figures to create course and distance pace averages. I have done this by adding up the pace scores of all the winners at a particular course and dividing it by the total number of races. The higher the average score, the more biased the course and distance is to horses that lead early or race close to the pace. Here are the 5 furlong handicap C&D pace averages for all turf courses in the UK.

 

Course 5f pace average 5f Pace Rank
Lingfield (turf) 3.33 1
Chester 3.3 2
Epsom 3 3
Catterick 2.97 4
Ripon 2.97 5
Redcar 2.88 6
Chepstow 2.86 7
Hamilton 2.85 8
Nottingham 2.84 9
Thirsk 2.82 10
Windsor 2.78 11
Musselburgh 2.77 12
Newbury 2.73 13
Beverley 2.72 14
Leicester 2.72 15
Pontefract 2.69 16
Goodwood 2.64 17
Ayr 2.63 18
Newmarket 2.58 19
Haydock 2.57 20
Wetherby 2.56 21
Bath 2.54 22
Doncaster 2.51 23
Salisbury 2.5 24
Sandown 2.5 25
Brighton 2.49 26
Carlisle 2.49 27
York 2.47 28
Ffos Las 2.38 29
Yarmouth 2.24 30
Ascot 2.24 31

 

Lingfield (turf) tops the list, but in truth they have very few 5f handicaps so we perhaps out to take this figure with the proverbial pinch of salt. Chester comes next which is no surprise based on the stats from the previous article. In that article Chester had exceptional winning percentages for front runners and very poor percentages for hold up horses. A 3.3 C&D pace average is huge, so let us look at Chester 5f in more detail.

Running style

Chester 5f

Wins Runners Strike rate (%) IV
Led 31 88 35.23 3.38
Prominent 21 194 10.82 1.04
Mid Division 5 109 4.59 0.44
Held Up 4 194 2.06 0.20

 

As can be seen, 52 of 61 Chester races have been won by horses that have either led or raced prominently. Essentially these figures indicate that the winner is almost six times more likely to be racing in the front half of the pack early on, than the back half.

Epsom are third on the list but they have only had 25 races so, as with Lingfield turf, the data is limited. Let us instead look at the Catterick who lie fourth on the list. Catterick have had 145 races so a bigger sample to breakdown:

 

Running style

Catterick 5f

Wins Runners Strike rate (%) IV
Led 47 196 23.98 2.51
Prominent 65 672 9.67 1.00
Mid Division 15 175 8.57 0.93
Held Up 18 473 3.81 0.4

 

The stats for Catterick are not in Chester’s league in terms of pace bias to front/prominent racers, but the tendency is still strong. Front runners especially have a very potent edge. Digging deeper, if we focus on races at Catterick with 12 to 14 runners the pace bias does increase significantly:

 

Running style Wins Runners Strike rate (%) IV
Led 15 66 22.73 2.88
Prominent 22 227 9.69 1.23
Mid Division 5 88 5.68 0.72
Held Up 4 201 1.99 0.25

 

37 of 46 races were won by early leaders or horses that raced prominent early. The winner is roughly four more times more likely to be racing in the front half of the pack early on, than the back half.

 

At this juncture I decided to dig a little deeper looking to see whether the going made a difference to the overall 5f course pace averages. In the past I have heard two contrasting theories connected with front running horses which would potentially affect the course pace average on a specific type of going:

Your first 30 days for just £1

Theory 1 – horses that lead on softer ground are difficult to peg back because horses find it harder to accelerate from off the pace on such going;

Theory 2 – horses that lead on firmer ground are likely to get less tired at the front due the faster conditions and this accentuates their front running edge. (Plus on quicker ground the race is likely to be run in a shorter overall time again meaning the front runner is expending less energy).

So which one is true – or is neither true? If front runners do have a bigger edge under certain going conditions it will push up the overall course pace average.

I decided to split the results into two – races on good or firmer; and races on good to soft or softer. Here are the course pace averages for all 5f handicaps split into these going types:

 

Going Course Pace average
Good or firmer 2.72
Good to soft or softer 2.67

 

As we can see the difference is minimal and not statistically significant. I plan to look at more extremes of going when I have time – looking at soft or heavy versus good to firm or firmer. However, looking at these initial figures, I am not expecting to see a huge variance.

My final area of research in this article is concerned with ‘class’. There is an argument, which I believe is a fair one, that the higher the class, the harder it is for horses to lead from start to finish – due to the more competitive nature of the opposition. Hence, at courses that run more higher class handicaps one might expect their course pace averages to be lower as a result. How to calculate ‘class’ at a particular course is difficult – do you use class levels, prize money, average Official Ratings across all races? I have decided to use a relatively simplistic approach by creating average class levels for each course by adding the class levels for each race and dividing by how many races there were. Hence, for example, if a course had had 10 class 2 handicaps and 10 class 3 handicaps their class average would be 2.5. Here are the course class averages for 5f handicaps (lowest class averages at the top):

Course Course Race Class Average Course Class Rank
Chepstow 5.47 1
Hamilton 5.43 2
Catterick 5.32 3
Brighton 5.26 4
Ffos Las 5.12 5
Beverley 5.11 6
Yarmouth 5.08 7
Bath 5.03 8
Carlisle 5 9
Nottingham 4.96 10
Redcar 4.95 11
Lingfield (turf) 4.92 12
Musselburgh 4.85 13
Ayr 4.77 14
Leicester 4.67 15
Ripon 4.57 16
Wetherby 4.56 17
Pontefract 4.53 18
Salisbury 4.45 19
Windsor 4.44 20
Thirsk 4.09 21
Goodwood 4.04 22
Newbury 4 23
Sandown 4 24
Doncaster 3.85 25
Haydock 3.79 26
Newmarket 3.64 27
Chester 3.02 28
Epsom 2.81 29
York 2.8 30
Ascot 2.62 31

 

As you would expect, most of the Grade 1 courses are near the bottom of the table. Three of these courses - Ascot, York and Epsom - have the most competitive 5f handicaps in terms of class.

To see if there is a correlation between course pace averages and average course race class I have ranked both lists next to each other, and produced an average rank. For there to be a strong correlation you would expect the majority of the courses to be in similar positions in each column – in other words the higher course 5f pace averages should correlate with the lower course class averages; likewise the lower course pace averages should correlate with the higher course class averages.

 

Course Course Class Rank (low>high) 5f Pace Rank Class / Pace Average
Catterick 3 4 3.5
Chepstow 1 7 4
Hamilton 2 8 5
Lingfield (turf) 12 1 6.5
Redcar 11 6 8.5
Nottingham 10 9 9.5
Beverley 6 14 10
Ripon 16 5 10.5
Musselburgh 13 12 12.5
Brighton 4 26 15
Bath 8 22 15
Leicester 15 15 15
Chester 28 2 15
Windsor 20 11 15.5
Thirsk 21 10 15.5
Ayr 14 18 16
Epsom 29 3 16
Ffos Las 5 29 17
Pontefract 18 16 17
Carlisle 9 27 18
Newbury 23 13 18
Yarmouth 7 30 18.5
Wetherby 17 21 19
Goodwood 22 17 19.5
Salisbury 19 24 21.5
Haydock 26 20 23
Newmarket 27 19 23
Doncaster 25 23 24
Sandown 24 25 24.5
York 30 28 29
Ascot 31 31 31

 

At both ends of the list, sorted by Class/Pace Average, we have the most valid correlations. For instance, Catterick, Chepstow and Hamilton all strongly favour front-runners and all host a majority of low grade five-furlong handicaps.

Meanwhile, Ascot and York, as well as to a lesser degree Sandown, Doncaster, Newmarket and Haydock, all generally host high class sprint handicaps where the early pace holds up less well.

I hope you have enjoyed this second instalment and, as always, comments are welcomed.

- Dave Renham

The Importance of Pace in 5f handicaps

This is my first article for www.geegeez.co.uk and before I start I would like to share with you my racing background, writes David Renham. I have worked for the Racing Post as a Spotlight writer and the Racing and Football Outlook as a trends ‘expert’; I have also written several books, mainly on draw bias, back in the early 2000s. And I have been a tipster with some success – and some failures! In all, I have written over 700 racing articles for magazines, newspapers, and websites.

Matt asked me to write on an ‘ad hoc’ basis which suits me as I have a full-time job outside racing at present. I hope you will find my articles interesting, useful, and ultimately lead to some profitable betting opportunities. However, as we all know, making money from backing or indeed laying horses is not easy. You need a combination of many things I believe – hard work; a good understanding of what you are trying to achieve; some sort of specialism as I feel there is simply too much racing and too many horses to gain a handle on if you don’t specialise; and, last but not least, a bit of luck.

For this article I am going to discuss pace in a race. When I talk about pace my main focus is the initial pace in a race and the position the horses take up early on. One of the many useful aspects of geegeez.co.uk is the pace section and the stats I am sharing with you in this article are based on the site’s pace data (found in the Pace tab on the racecard).

The pace data on Geegeez is split into four - Led, Prominent, Mid Division and Held Up. Let me try to explain what type of horse fits what type of pace profile:

Led – essentially horses that lead early, usually within the first furlong or so; or horses that dispute or fight for the early lead;

Prominent – horses that lay up close to the pace just behind the leader(s);

Mid Division – horses that race mid pack;

Held Up – horses that are held up at, or near the back of the field.

So after each race all the horses are assigned points in regards to what position they took up early in the race. Leaders get 4, prominent runners 3, horses that ran mid division 2, and those held up score 1. Geegeez has over 1,059,000 runners’ pace comments scored, from a total of about 1,100,000. [The others are things like unseated rider at the start, or where there is no discernible pace reference in the comment].

If you click the pace tab on the website you are presented with pace data regarding the specific course and distance of that race, and pace data for each horse covering their last four UK or Irish runs. For this article I am concentrating on the course data and creating pace figures for specific course and distances – namely handicap races run over 5 furlongs. I have always been a fan of sprint handicaps and early pace in sprint handicaps generally gives a bigger advantage to front runners than races over longer distances. In addition to this, some courses offer a bigger advantage to front runners than others as you will see.

The first set of data I wish to share with you is the overall pace stats for 5f turf handicaps (minimum number of runners in a race 6):

Pace comment Runners Wins SR%
Led 3450 637 18.5
Prominent 9987 1078 10.8
Mid Division 3187 235 7.4
Held Up 8465 567 6.7

Horses that led, or disputed the lead early, have a huge advantage in turf 5f handicaps. So, if we could predict the front runner or front runners in each race we should be ‘quids in’, and indeed would be. Unfortunately, it is not an exact science and how best to do this I will leave for a future article.

Best performing 5f handicap tracks for front runners

My aim for this article is to show you the differences in the course figures for 5f handicaps and how some courses are more suited to early leaders/front runners than others. Here are the courses with the best strike rates (minimum 40 runners):

Course Front Runners Wins SR% P/L SP IV
Chester 88 31 35.2 120 3.38
Catterick 196 47 24 177.71 2.51
Hamilton 170 39 22.9 130.29 2.04
Beverley 197 44 22.3 167.29 2.51
Epsom 50 11 22 45.5 2.96
Nottingham 219 48 21.9 224.08 2.32
Leicester 88 19 21.6 60.75 1.91
Windsor 160 34 21.3 100.31 1.9

 

Chester has amazing stats for early leaders: the tight turning 5f clearly suits front runners and, when combined with a good draw, front runners are clearly hard to peg back. Another round 5f, Catterick lies second with excellent figures also. Keep in mind that the average strike rate is 18.5% for all courses over this minimum trip.

Worst performing 5f handicap tracks for front runners

At the other end of the scale here are the courses with the poorest stats for early leaders/front runners in 5f handicaps:

Course Front Runners Wins SR% P/L SP IV
Newmarket (July/Rowley combined) 88 12 13.6 -8.37 1.19
York 106 14 13.2 21 1.78
Haydock 146 18 12.3 -18.17 1.25
Sandown 119 13 10.9 -19.37 1.04
Yarmouth 96 10 10.4 -39.58 0.86
Ascot 98 8 8.2 -30.5 0.99
Doncaster 90 6 6.7 -32.5 0.81

 

It is interesting to see York in this list – York is often considered a decent front running track, but not according to our figures.

 

Chester performance by number of runners in race

Looking at Chester in more detail, we can split the data by number of runners:

Runners in race Front Runners Wins SR% P/L SP IV
6 to 8 36 18 50 90.5 3.65
9 to 11 35 11 31.4 23.5 3.22
12 to 14 17 2 11.8 6 1.46

 

Your first 30 days for just £1

Here at geegeez.co.uk, data regarding number of race runners is calibrated slightly differently to my table, but you are able to change the figures on the site to suit your own personal requirements.

 

Overall performance by number of runners in race

As we can see from the Chester figures, the smaller the field size, the better it has been for front runners. The general perception of punters I believe matches the Chester data – in other words most punters believe front runners are more likely to win in smaller fields. It makes sense I guess as there are less rivals to pass the leader. However, is this really the case? Here are the data:

 

Runners in race Front Runners Wins SR%
6 to 8 1214 264 21.7
9 to 11 1205 223 18.5
12 to 14 624 106 17.0
15+ 407 44 10.8

 

The stats back up the basic theory, but a 17% win rate for early leaders/front runners in 12 to 14 runner 5f turf handicaps is a strong performance, especially when you take into account the likely prices of such runners. Hence, one could legitimately argue that the best front running value lies in the 12-14 runner range.

 

Best performing 5f handicap tracks for hold up horses

Of course, early leader/front runner stats are not the whole story when trying to build up a ‘pace’ picture of each course. We need to look at the stats at the other end of scale – those for hold up horses. Firstly a look at the 5f courses that offer hold up horses the best strike rates:

Course Hold up horses Wins SR% P/L SP IV
Yarmouth 195 27 13.8 -33.04 1.16
Bath 332 41 12.3 -9.5 1.1
Brighton 258 30 11.6 -68.97 0.89
Newbury 99 9 9.1 -31.92 0.82
Salisbury 66 6 9.1 -23.5 0.8
Leicester 178 16 9 -51.87 0.79
Carlisle 192 17 8.9 -55.25 0.82

 

Interestingly you would expect these courses to match those that have the poorest stats for early leaders/front runners (see above). However, only Yarmouth appears in both groups. Hence the importance of not just looking at the ‘led’ data in order to appreciate pace biases at particular courses.

More materially, perhaps, all courses are firmly negative at SP, and most have an impact value of less than 1, meaning such types are less likely than horses with other run styles (1 meaning the same likelihood).

Worst performing 5f handicap tracks for hold up horses

Now a look at those courses with the worst strike rates for hold up horses:

Course Hold up horses Wins SR% P/L SP IV
Chepstow 187 10 5.3 -104.42 0.5
Musselburgh 746 39 5.2 -346.17 0.5
Ripon 200 8 4 -122.42 0.38
Redcar 307 12 3.9 -200.92 0.41
Catterick 473 18 3.8 -312.17 0.4
Epsom 113 3 2.7 -98.25 0.36
Chester 194 4 2.1 -160.5 0.2

 

Chester, Catterick and Epsom appear in this table – courses that appeared in the top 5 for front runners. However, once again the correlation between good courses for front runners / poor courses for hold up horses is not as strong as one might expect.

What can be said with a degree of confidence is that these tracks are graveyards for hold up horses and such runners make abject bets in the main.

Summing Up

So how should we use the data discussed in this article? There are numerous ways to do this, some of which I will elaborate upon in a future article. Ultimately however, it is important to appreciate the differences between each course and distance in 5f handicaps, especially their configuration and favoured run styles, points which should inform your betting when you decide to use pace data as part of your betting strategy.

For example, if you feel you have found two ‘nailed on’ front runners in two different 5f handicaps, at say Chester and Yarmouth, you need to appreciate that whoever front runs in the Chester race, has, according to past data, over 3 times more chance of winning than your Yarmouth trailblazer. Of course your ‘nailed on’ front runner might not lead early but that is not really the point I am trying to make!

I hope you have found this article interesting and potentially useful from a betting perspective. If you have yet to use the pace data on geegeez.co.uk, I hope I have sown some seeds of interest and that you may start to think about how to incorporate pace handicapping into your betting armoury.

- David Renham

** You can read Part 2 of this series here **

Easter Eggs from Geegeez…

It's Easter so how about some Easter eggs? In the software community, an Easter egg is a hidden piece of functionality only known to insiders. Well, we're not quite so covert here at geegeez.co.uk, so in today's post, I've a couple of tasty morsels which are hopefully better for you - for your teeth at least - than a chocolate ovoid!

Specifically, I've recorded a video showing what's new and what's coming soon, including something about which I'm very excited. And I've a few words - mainly just for fun - on the Irish National. Let's start with the Geegeez Gold Easter Egg...

Now and Next on Geegeez Gold

Video timeline

00:00 Introductory waffle

00:30 Hcap/All filter on Draw tab

03:08 The first change to Instant Expert for a loooooong time

05:17 Query Tool: cosmetic enhancements

07:38 Query Tool: pictures!

08:55 Query Tool: Angles, a first look

10:25 QT Angles on the racecard

14:30 Query Tool: example Angle

19:10 Two-day and one-week Gold passes (non-recurring)

**

Irish National Thoughts

The following is at best half-baked, and is offered 'just for fun' on what is shaping up to be something of a washout day...

I would be slightly sweeter on the picks were it not for the white hot tussle for the Irish Trainers' Championship between Messrs Mullins and Elliott. That significant set to has seen no fewer than THIRTEEN entries from Cullentra House, and another four from Mullins' Closutton base. Whatever you think of that from a macro perspective, it makes for a very different race shape and, thus, trends - which have been strong in this contest down the years - may be torpedoed on this occasion.

Your first 30 days for just £1

No matter, for the trends have been strong, as I say. A low weight, and a commensurate lesser rating; a younger horse; a win over a trip of three miles or beyond; fairly unexposed over fences; and a recent run between two weeks and two months. Those criteria (perhaps excluding the weight/rating one) largely fit a lot of top table staying chases and so make sense.

In truth, it is the weight/rating element which I suspect may derail us this term, with so many classy contenders from the two aforementioned powerhouses. One thing you can be sure if you fancy one from Mullins' or Elliott's batallion is that they will have no idea which is most likely to win: jockey bookings and the like may be less of a pointer then, and prices may be artificially inflated down the batting order.

Enough of the flim flam, here are the ones I've backed:

Trends picks

My trendy shortlist was Arkwrisht, Squouateur, Sutton Manor and Champagne Harmony. The last named is a million chance on form and will take no more than the minimum stake on the machine as a flippant action bet. I have reservations about both S's, Squouateur and Sutton Manor, in a big field, though I've backed them both, at 25 and 40 on Betfair - both are now bigger (great, ahem).

The one I like most is Arkwisht, trained by Joseph O'Brien, and ridden by Rachael Blackmore.

Surprisingly, Elliott and Mullins are collectively nought from 50! That must be odds-on to change this afternoon, but if not Arkwrisht ticks a lot of boxes - form in big fields, on heavy ground, and beyond three miles - though I do have a slight niggle about whether the eight-year-old German-bred can quite see out this 29 furlong slog. Still, each way at 28/1 with as many place concessions as you can snaffle should offer a run.

Other picks

And I've taken a couple of other 'just for fun' swipes - well, you can't have enough swipes in a race like this, can you? - in the form of The Paparazzi Kid and Moulin A Vent.

The Paparazzi Kid would be a real story horse if he won. €270,000, the first prize, could well be more than the difference between Mullins and Elliott at the end of the season, so the fact that this chap is having his first start for Gordon, having previously been trained by Willie, would smart in the extreme for the incumbent champ!

He has a better form chance than 66/1 (160 on the exchange) implies as well, in spite of being a full on trends buster. He's eleven (too old), his winning form is all at two and a half miles (won't stay), and he's been chasing for five years (too long). The negatives out of the way, the case for the defence is thus: loves heavy ground (122114 lifetime), has very good form in big fields (20311244U in 16+ runner races), has slipped to a pound below his last winning mark, and has gone to a yard with an excellent record at winning first time after a trainer switch and off a layoff.

He'll be ridden for luck, which he very well might not get, and to get the trip, which he very well might not get; so caveat emptor. But he'll be fun to watch for a circuit and a bit at least.

Moulin A Vent has a more obvious claim, as one of the classier horses in the line up. This Graded performer drops into handicap company for the first time, though jockey bookings suggest he's the stable second string, Sean Flanagan jumping ship to Snow Falcon. He may live to regret that as his spurned former partner, on which he's sat every time that one has raced over an obstacle, sets up well against conditions.

Only six, Moulin A Vent has had 14 career starts, just six of which have been in chases. He's clearly unexposed - the corollary to which is that he's inexperienced - and a fast run stamina test might be just what he wants. Big fields and heavy ground hold no fears, but the fences do: this chap is a sticky jumper. Faller insurance may be a prudent play if you can find it, and the general 33/1 - nearer 40's on Betfair (no insurance) - is attractive.

There are loads of class horses in what is a really deep and competitive renewal, so take a few stabs, swot up on your Gigginstown cap colours, and strap yourself in for a fun ten minutes or so. Good luck!

Matt

More New Bits on Gold: Pace Granularity

As promised, we've added a couple more new bits to Geegeez Gold - with plenty more to follow in the coming weeks.

Today, I'm pleased to share with you improved pace granularity and the addition of the HCAP/ALL filter on Report Angles. Let's deal with the latter first.

Report Angles: Handicap Only option

On a number of Gold reports, including Trainer and Jockey Statistics, Trainer/Jockey Combo, and Trainer and Sire Snippets, it is possible for a user to select the data based on ALL races or handicaps (HCAP) only. Well, by surprisingly (to me, at least!) popular demand, we've added these filters to the new Report Angles feature.

PLEASE NOTE: We're aware of a problem with the Trainer Snippets HCAP options on Report Angles, and working to fix that. For now, please leave them set to ALL if you you use Trainer Snippets within Report Angles.

It looks like this:

On selected reports, you can now opt to view Angles data for handicaps only, or for all races

On selected reports, you can now opt to view Angles data for handicaps only, or for all races

 

Once you've set the angles up - don't forget to save them - you'll be able to see your chosen parameters both in your Report Angles report, and on the racecards themselves, as follows:

If you've selected HCAP on the settings page, you'll only see handicap race data on your Report Angles

If you've selected HCAP on the settings page, you'll only see handicap race data on your Report Angles

 

Report Angles appear inline on the selected race types

Report Angles appear inline on the selected race types

Your first 30 days for just £1

 

Pace Granularity

Our pace information is incredibly instructive for understanding how a race will be run. The pace maps are rarely far from what comes to pass and are a must for the serious punter. But, when it comes to understanding how the shape of today's race overlays onto history, we had hitherto adopted a 'one size fits all' approach.

That is, we lumped all races over a given course and distance combination together, regardless of whether there were five or fifteen runners; or the going was firm or heavy. Clearly that's a little too imprecise to be optimal, so we've addressed it.

From today, we have implemented going and field size ranges to better capture today's race shape against its historical precursors. The short video below explains all:

 

 

There is plenty more in the pipeline, and I look forward to sharing it with you soon.

Matt

p.s. if you have any issues with getting the new elements to work, please do consult this FAQ before contacting our support. Obviously, if you've done that and are still having problems, let us know!

New Gold Features Coming Soon in 2018…

New Year, New Year.

Happy New Year to you.

After a quiet spell, we're ready to kick things up a notch both in terms of editorial and our premium racing form provision. In today's post, I'd like to share with you what's coming next to Geegeez Gold...

Right Now

We made a couple of very small changes yesterday to:
- include official ratings on full results
- publish the winning time on full results
- add Report Angles to the report dropdown

Here's how those all look:

 

In the next fortnight

Meanwhile, being put through its paces on test as I speak are a couple of more meaningful changes.

HCAP option on Report Angles

The first of them is the addition of a HCAP option to Report Angles. When we released Report Angles a month or so ago, the first response was 'Wow!' - after we got over a couple of teething issues, that is. The second response was, "Can you add a handicap only filter, please?".

Well, we're here to bring you the stuff you want, so yes, we heard you and I'm happy to say this will be live in the near future. It's already up on my test setup, as you can see from the below image:

We've added 'Handicap only' filters to Report Angles in line with the individual reports

We've added 'Handicap only' filters to Report Angles in line with the individual reports

 

Pace tab: enhanced focus

The second enhancement is to the Pace tab. "In the land of the blind, the one eyed man is king" has been our motto on this front to date. That is to say that, because nobody else is doing anything meaningful on pace (which, incidentally, is the singlemost under-rated element of form study in this country, in my view), our Pace content stands apart at making it easy to see how a race will likely pan out.

But... it's been an irritant to me for some time that, for instance, the historical pace profile of a big field fast ground race may very well not be the same as for a small field soft ground race over the same course and distance.

So we've addressed that, by adding going and field size dropdowns to the Pace tab.

This is Chelmsford's 8.15 race tonight, a 7f contest between eight runners. As you can see from the top part of the image - with all going and all field sizes selected - it has been advantageous to be on the lead. Indeed, 21% of winners have raced that way for a profit of £34 to a £1 level stake.

Your first 30 days for just £1

But look at the bottom part, where we focus on standard going (which is most of the runs, of course) but also on a more targeted runner range of seven to ten. Now we can see that those horse which led have a 28% win rate and a £1 level stake profit of £73.71.

Hopefully you can see how this granularity will be useful. I'm very excited about sharing it!

[One caveat is that it will often produce small sample sizes. That's why we have the dropdowns so users can extend the ranges to get a more meaningful sample.]

Field size has a major bearing on run style effectiveness

Field size has a major bearing on run style effectiveness

 

**

Next six weeks or so

A little further down the line we have still more planned. I don't want to go into too much detail yet, mainly because I haven't got screenshots to show you.

But there are a couple of extra reports (class move, and SR differential) which will be interesting.

We're also looking at an Instant Expert 2.0, which will allow users to filter by Handicap/All Races, and by different time periods.

And we'll be creating some filters on the racecard menu, so you can see those races which interest you. Not interest in small field races? Only want handicaps to display? Want to exclude Irish, or UK racing? Only interested in sprints? Just want the better class stuff? You'll be able to filter by all of these things, and have your racecard menu presented to you... very soon.

Next three to six months

Looking further down the line we're working on making Geegeez Gold less prescriptive and more in line with what you want to see. We'll be enabling you to create and save angles in the Query Tool; and you'll be able to see angle qualifiers right within the racecard.

For those short on time, we're also working on a Bet Finder feature. This will identify horses running today that meet certain fixed criteria outlined by you.

Lots of exciting developments in the pipeline, and I hope they'll provide you with even more of an edge over your fellow punters.

Matt

New to Gold: Report Angles

Today, I'm pleased to introduce you to the latest Geegeez Gold feature, Report Angles.

As part of our commitment to extend greater flexibility and configurability to Gold users - in plain English, to let you do more of what you want to do! - we've created an aggregator for all reports. You can set it up as you wish, or not at all if that's your wish.

More details are in this video, and in the article beneath.

N.B. All angles are turned OFF by default. Read/watch on to discover how to turn them ON.


Report Angles: Overview

 

Report Angles highlight content from Gold’s existing set of reports against today’s runners as displayed on the racecards.

That is, for each report, there are now – as of December 2017 – a group of pre-set parameters which, when matched, will be flagged against a runner on the racecard.

Using the example from above, the Trainer Statistics report might have the following pre-set parameters for its Type 1 (i.e. 14 Day Form) sub-report:

  • 10+ runs
  • 30%+ wins
  • A/E 1.25+

Where a runner satisfies those criteria, it is highlighted on the racecard as such. There will be pre-sets for every report sub-type, e.g. Trainer Stats report will have four pre-sets, one each for 14 Day, 30 Day, Course 1 Year, and Course 5 Year.

Users will be able to select any or all of the pre-sets to be displayed on their racecard views. They will also be able to edit or restore to default the pre-sets. However, a user may only have one custom view of each report sub-type.

Report Angles are automatically built into the ‘My Report Angles Settings’ page. Users have the ability to activate, deactivate, amend or restore to default each Report Angle. They cannot create new Report Angles, however.

 

Report Angles: My Settings

Users can select, de-select, amend and/or reset the Report Angles configuration on the My Report Angles Settings page. However, users cannot create or delete Report Angles, though they can disable/enable them.

The page is found at https://www.geegeez.co.uk/reports/my-report-angles-settings/ and looks like this:

The Report Angles Settings page displays the report titles (i.e. TJ Combo, etc) on the left-hand side, with settings displayed for the selected report sub-types (e.g. 14 Day, 30 Day, Course 365 Day, Course 5 Year).

For each report/type combination, there are editable parameters as per that report’s individual report page. For example, below are the editable parameters for four sub-types of TJ Combo report:

N.B. Different reports have different parameters – users are advised to check each one individually, at least the first time they configure the settings.

At the bottom of the screen are three blue bars. The first, “Save Settings”, enables a user to save any changes made within the selected report.

The second, “Reset Defaults”, reverts the selected report to the ‘factory settings’. The third, “Reset All Defaults”, reverts all reports and sub-types back to their default settings.

N.B. These defaults are NOT optimal. Rather, they are presented as a balance between limited data and too much data appearing in the report. Users are encouraged to experiment with the settings to find the appropriate volume of report output.

Your first 30 days for just £1

Each report sub-type has a tick box next to its name. Selecting/de-selecting the ticks will include/exclude a sub-type from the report and racecard view.

Clicking the on/off buttons top RIGHT will select/de-select all tick boxes for a report.

Clicking the on/off buttons top LEFT will select/de-select all tick boxes for ALL reports.

 

Report Angles: The Report

Once a user has selected and/or activated report angles and parameters, all qualifying runners will appear on a report on the My Report Angles screen. The report looks like this:

 

Each row in the report table is clickable, and will open the race in question in a new window. All columns are sortable to enable users to configure the view to suit personal taste.

 

Report Angles: Racecard Inline

The racecard has been updated with a new ‘report’ icon, containing a numerical indication of the number of angles matched. Clicking the icon will reveal inline the qualifying Report Angles, as in the below example.

There is also a new icon with a ? in the top icon menu. Clicking this icon will open Report Angles in the card for all runners. Clicking again will close them.

 

Getting Started with Report Angles

By default, all Report Angles are switched off. To turn them all on, use the 'ALL On' button top left on the Report Angles Settings page. Alternatively, and preferably, take a few minutes to set the Angles up as you would have them.

The default settings, when all Report Angles are switched on, can be seen in the below table.

 

Report Angles are intended as an aid to successful betting; they are not to be used as an end in themselves. That is to say, Report Angles may highlight interesting elements about certain runners but, as with all other approaches, a more holistic consideration of the puzzle will always yield better results.

Good luck, and I hope you enjoy this new feature as much as I have been during the testing stage.

Matt

Winter Gold: Day 1

In the first of five special features looking in detail at Geegeez Gold this week, I introduce an overview of the site. It includes tips, reports, tools, and a couple of very smart under-the-radar trainers to follow!

Click the video to start watching... (and 'square' icon bottom right of video to go full screen)


SPECIAL OFFER: SIX MONTHS GOLD, ONE TIME PAYMENT (NO RECURRING CHARGES)

Saturday Racing Video Preview

After Friday's report dissection - good conceptual intel, awful results! - I'm back with another video blast for Saturday's Ascot card. It's a really decent meeting, and I'll probably have retire in disgrace if I draw a blank. Gulp.

Anyway, with hopefully some more 'how to' insight as well as perhaps a chicken dinner or two, here's the video.

As with yesterday, if you have audio problems, click this link to watch it on the vid hosting site.

Good luck!

Matt

Your first 30 days for just £1